System Design Fundamentals

0% completed

​

Trade-offs in CAP Theorem

Now let's discuss the heart of CAP: the trade-offs. Why can’t a distributed system have Consistency, Availability, and Partition tolerance all at once? The simple answer: when a network partition happens, the system has to make a tough choice:

  • If the system decides to remain consistent (C), it cannot allow conflicting updates on different sides of the partition. So at least one side of the partition must stop accepting operations (to avoid divergence). That means some part of the system becomes unavailable (not serving requests)

.....

.....

.....

Like the course? Get enrolled and start learning!
manoranjan das

manoranjan das

· 3 months ago

When we claim CA is possible only when there is systems with out any partition meaning no nodes or systems talking over n/w then how it can be A. As we need redundant systems for A as per definition. So in that case how does it work ? Seems A and P always together otherwise there is no A.

This part: "No matter which node you contact, if it hasn’t crashed, it will respond within a reasonable time. An available system never resorts to saying “sorry, I can’t serve your request right now” due to a node or network failure.

In practical terms, this often means there are multiple redundant nodes, so if one goes down or can’t be reached, another can serve the request."